Quality Assurance Mechanisms in Agrifood

38 Pages·2004·0.43 MB·English

CORE Metadata, citation and similar papers at core.ac.uk
Provided by Research Papers in Economics
Quality Assurance Mechanisms in Agrifood:
The Case of the Spanish Fresh Meat Sector∗
Manuel González-Díaz**, Marta Fernández Barcala** and Benito Arruñada***

Published in a joint special issue of the International Journal of Technology
Management and the International Journal of Agricultural Resources
Governance and Ecology, 2 (3/4), 2003, 361-82.

First version: September, 2001
This version: March, 2002

Abstract
The largest fresh meat brand names in Spain are analyzed here to study how quality is
signaled in agribusiness and how the underlying quality-assurance organizations work.
Results show, first, that organizational form varies according to the specialization of the
brand name. Publicly-controlled brand names are grounded on market contracting with
individual producers, providing stronger incentives. In contrast, private brands rely
more on hierarchy, taking advantage of its superiority in solving specific coordination
problems. Second, the seemingly redundant coexistence of several quality indicators for
a given product is explained in efficiency terms. Multiple brands are shown to be
complementary, given their specialization in guaranteeing different attributes of the
product.

Key words: quality assurance, co-branding, agriculture, vertical integration, contracts.

∗ The authors thank the assistance of Eduardo Melero, Neus Palomeras, Roberto Sánchez and
Luis Vázquez in gathering part of the information; the comments of Demian Castillo and Luis
Vázquez; and the financial support from the European Union and the CICYT, a research agency
of the Spanish Government, through the FAIR PL98-4404, and the SEC-99-1191 and
SEC2002-04471-C02-02 projects, respectively. Usual disclaimers apply.
** Department of Business Administration. Universidad de Oviedo. Avda. del Cristo s/n,
33071-Oviedo, Spain. E-mails: [email protected]; [email protected].
*** Department of Economics and Business. Universitat Pompeu Fabra. Ramon Trias Fargas,
25-27, 08005-Barcelona, Spain. E-mail: [email protected].
JEL codes: D23, L14, L22, Q13.
1. Introduction
Brand names have become surprisingly important in the fresh meat market. Fresh meat
has traditionally been sold mainly without brand name. However, 5-10% out of the
fresh meat is now sold as quality meat.1 In this, up to three simultaneous brand names
are used, referring meat’s origin and quality. It seems, therefore, that brands have
replaced the traditional butcher’s role in guaranteeing meat quality with his own
reputation. This is probably due to two causes. First, the increase in the opportunity cost
of the traditional housewife has changed shopping habits from the traditional butcher
model towards the supermarket and hypermarket model.2 Second, health scares such as
the “BSE crisis”, the “Belgium chickens” or the “foot-and-mouth disease fever” are just
some of the terms which European consumers associate with fraud in meat products in
the last five years. In fact, although total fresh meat consumption has increased by 1.4%
from 1994 to 1999, beef and chicken consumption have decreased in 9% and 7%,
respectively.3 Consequently, brand names have become the main quality signal for
consumers, motivating the establishment of numerous new brands.
Those brand names differ substantially among them. On one side, they signal different
dimensions of quality. In fact, some of them do not indicate “high quality” attributes,

1 Internal estimations of Empresa Asturiana de Servicios Agrarios for 1999.
2 In 1994, purchases of fresh meat in major retail outlets (supermarkets and hypermarkets) were
at 47.85%, rising to 49.19% in 1999. Likewise, shopping in traditional shops decreased in the
same period from 41.26% to 38.27% (MAPYA, 2000, p. 207). The same tendencies have been
maintained since the eighties.
3 MAPYA (2000, pp. 59-60).
1
but only the homogeneity of their products. In this sense, it is not difficult to find
reputable brand names in which awareness comes mainly from the homogeneity of the
product instead of its organoleptic attributes (McDonalds’ hamburgers could be an
example). On the other side we can differentiate two types of brand names, depending
on who guarantees the quality. First, when the guarantee comes from a private firm, the
quality signal is the typical private brand name. For example, in agribusiness, it is
frequent that the distribution companies, usually the larger ones, launch their own
brands, guaranteeing product quality themselves (such as the Spanish “Corte Inglés” or
the French “Carrefour”).4 A variation of this type of brand name comes about when a
well-reputed company is hired to guarantee consumers the quality of other product. This
is the case of the of quality certificates, such as those of Lloyds or AENOR, which
guarantee to the consumer that the producer follows certain rules and procedures to
detect and avoid bad quality output. Second, a public or governmental institution can
guarantee consumers the quality of the product. It is when a company sells its product
sheltered by the prestige of some specific geographical spheres and/or production
methods related to a superior product quality. In this case, we refer to this quality signal
as “Geographical Indicators” from here on out, including Protected Denomination of
Origin (PDO),5 Protected Geographical Indicator (PGI) and the Guarantee Brand Name.
Although there are some differences among them regarding the degree of exclusiveness
within a territory, all are brand names protected by the European Union.6 Both

4 The owner can also adopt other legal forms, such as co-operatives or associations of
producers. In those hybrid cases the only difference lies in the collective ownership of the brand
name that guarantee the quality to the consumers.
5 See on this, Castillo (2002).
6 The process undergone by a hypothetical product with denomination could be as follows: an
Autonomous Region grants the Guarantee Brand Name. The product is therefore accepted
2
categories of brand names, private and public, are however nonexclusive, overlapping
subsets: we observe that one single product could be retailed with more than one brand
name.
The aim of this paper is twofold. On one hand we want to analyze which mechanisms of
governance are used to solve different dimensions of the information asymmetry (high
attributes and their homogeneity). On the other hand, we want to explain the economic
rationale of “co-branding”: the concurrence of more than one brand name on the same
product.7
The remaining of the article is organized as follows. First, we describe the typical
problems and solutions involved in guaranteeing the quality of products. Section three
defines our two hypotheses: First, the mechanisms of governance chosen to manage a
brand name will depend on the objective of each brand. Second, the different types of
brands that we observe are complementary in assuring quality. These hypotheses are
tested using six case studies, which are described in section four. The empirical test is
discussed in section five. Section six concludes.

within the regional territory. As long as the product is suitable, the Autonomous Region
proposes it as a Specific Denomination or as Denomination of Origin, depending on the nature
of the product, being confirmed by the Ministerio de Agricultura, Pesca y Alimentación
(Spanish Ministry of Agriculture). In this case, the product is accepted at both the national and
international levels. To be a product with a Specific Denomination or Denomination of Origin,
the product does not have to have first been a product with a Guarantee Brand Name. The
products recognized by the Spanish nation are also recognized by the EU as products with a
Protected Geographical Indicator, Protected Denomination of Origin or Guaranteed Traditional
Specialty.
7 Co-branding is not exclusive of agrifood products. Credit cards and computers are other well-
known examples. See Blackett and Boad (1999) for an extensive overview on this topic.
3
2. Quality Assurance
2.1. The Problem
The problem of quality lies in the fact that the consumer cannot always determine the
quality level of a product before purchasing it. This information asymmetry increases
the chances of opportunism,8 which can even prevent the transaction from taking place.9
The intensity of this problem depends on the characteristics of the product. On this
subject, three types of product attributes that determine their potential controversial
nature have been identified: search, experience and confidence,10 of which only search
attributes can be determined before purchasing. The larger the influence of experience
and confidence attributes into the consumer utility function, the bigger will be the
information asymmetry problem and the more interested the producer will be in
assuring the product’s real quality.
On the other hand, the information asymmetry of a product’s quality presents two
dimensions: the average or expected quality of a producer or brand and the deviation of
each product from that average value.
The first dimension refers to the qualities consumers may notice in the different
attributes that form the product and the way they value them. This is often called
“subjective” or “design” quality and is related to the degree to which they satisfy the

8 See, for example, Williamson (1985, pp. 47-48).
9 See, for example, Akerlof’s classic work (1970).
10 See Nelson (1970) and Darby and Karni (1973).
4
customer’s preferences.11 They usually coincide with the more limited and difficult-to-
obtain attributes, at least among regular or experienced consumers. Agrifood products
are no exception, and although it is commonly said that “there is no accounting for
taste”, consumer preferences are far from being determined by chance, usually
concentrating on well-known and defined organoleptic characteristics.
The second quality dimension refers to the homogeneity between products of the same
producer, or protected by the same brand. This dimension is related to the degree to
which the pre-established design conditions are observed and is often called “objective”
or “conformance” quality.12 In the agrifood sector, these aspects were not traditionally
of concern, and were more typical of industrial products than of agricultural or cattle
products. However, the success of products considered as low-quality at first but very
homogeneous—such as hamburgers and sausages—shows the importance that the
consumer of foodstuffs gives to being sure about the quality of the products purchased,
regardless of their organoleptic attributes. Nevertheless, this difficulty has a
technological origin since it is hard to know how the different variables affect the final
quality of the product. This makes homogeneity one of the main concerns for producers.
2.2. The Solution
The most common way of solving the problems caused by informational asymmetry on
quality is that the informed party (the producer) signals its private information by
adopting a behavior (by investing in brand name) that, properly interpreted, reveals their
information to the non informed party (the consumer). This signaling involves

11 See Edwards (1968).
5
committing the quasi-rent from his reputational capital in every exchange.13 In this
sense, the producer must first create a reputation for his brand name in order to be able
to use it later as a guarantee. For that purpose, he must start offering high quality in
repeated transactions. Only in that way, after repeated purchases of experimented or
confidence goods, will consumers gradually realize that the quality offered is suitable
and consistent with the passing of time, being confident that they will not be deceived.
In exchange for that superior quality, consumers are willing to gradually pay for that
brand name a price increase or “premium” with regard to other products that have not
been signaled and do not offer these guarantees.
In perfect competitive markets, this premium represent the normal profitability of the
investment made by the company in its reputational capital, and will depend on the
problems it resolves for the consumer. Thus, it is likely that the consumer dislikes both
a low average quality of the products and a high deviation from expected or mean
quality. Thus, a high average quality would imply that the consumer is willing to pay a
positive “quality premium”, and a low variance from that mean would cause a
“homogeneity premium” (see Table 1).

12 See Crosby (1979, p.15).
13 See Klein and Leffler (1981) and Shapiro (1983) for two pioneer models of this process.
6
Table 1: Quality Premiums in Agrifood Products
Subjective quality (organoleptic attributes)

superior Standard
Ideal situation: “quality premium” “Homogeneity premium” (but no
High
plus “homogeneity premium” “quality premium”)
Objective quality
(homogeneity)
The worst situation: neither “quality
“Quality premium” (but no
Low premium” nor “homogeneity
“homogeneity premium”)
premium”

The combination of quality premiums, the specificity of reputational capital and
repeated transactions make the producer's signaling credible. Consumers know that if
they are deceived, the present value of the company’s reputational capital decreases
because its future transactions would be endangered if their clients do not trust them.
The business of the producer is not to take advantage in a short-run, but to obtain
“normal” profits from his investments from many long-run exchanges. In other words,
the producer would lose the value of his quasi-rents, generated from the investments
made to create his reputation.
3. Hypotheses
3.1. Specialization and Contract Design
Our first hypothesis argues that depending on the objective of each brand name in
solving quality problems (low mean quality and product heterogeneity), the mechanisms
7
of governance selected and the contract design will be different. Brands that place more
emphasis on the homogeneity of the product (i.e. they are seeking to reduce the variance
of the attributes of the product) display governance mechanisms similar to hierarchies.
Conversely, those that attempt to guarantee mainly the average characteristics of the
brand are more likely to be based on the market.
The theoretical justification for this argument is based on the conjecture that some
mechanisms are more appropriate for solving coordination problems (hierarchies) while
others are for solving motivation problems (markets). Where the aim is to obtain a
homogeneous product, the chief goal is a good coordination of inputs, even more than
having a perfectly designed product or the avoidance of small local imbalances in the
inputs. For this kind of decision, in which the main problem is the timely
synchronization of activities and resources, transaction cost literature argues that it is
more costly to coordinate activities through the market than through hierarchy.14 On one
hand, probably the best information about coordination is only available to the party that
is central to all the other agents involved. However, in order to make the market work, it
would be necessary to transmit all that key information to each decision-maker and this
would entail important transfer costs. On the other hand, it is also feasible that the
reaction of each agent to changes in prices would depend on their wealth and risk
preferences, hindering the coordination goal.
The market offers better results when the characteristics sought are not so much the
coordination and homogeneity as the superiority of the specific organoleptic attributes.
In this case, the relevant information (how to obtain the best attributes) does not come

14 See Williamson (1985 and 1991) and Milgrom and Roberts (1990 and 1992, pp. 91-92).
8
from this central party, but is dispersed throughout the organization or even among the
different producers (the specific know-how of each producer). Moreover, in this case,
according to the theory of incentives, it is possible to offer internal incentives to the
different agents so that they feel appropriately motivated to seek these attributes. Market
transactions usually provide this kind of incentive.
We can also anticipate the characteristics of the contracts used in these two types of
transactions. We expect more complete contracts in those brands that aim mainly at
homogenous products, since it is necessary that the performance of each agent be
precise and coordinated. This requires detailed performance guidelines for the parties,
whether they are employees or external business associates. On the contrary, when the
objective of the brand is a high subjective quality based on organoleptic attributes, it is
common to use relatively incomplete contracts, in which only general performance
guidelines and the mechanisms for solving problems are established, without specifying
ex ante specific performances.
On the other hand, the decision-making capacity of the agents will be much higher for
those brands that seek the superiority of their organoleptic attributes than for those that
try to guarantee a homogeneous product. In the first ones, specific local knowledge is
fundamental in order to develop high-quality products. This is difficult to transmit given
the great variety of conditions each farming enterprise has to face. On the contrary,
when the basic aim is to guarantee the homogeneity of a product, all the agents must act
in coordination in order to offer identical answers to the same problems.
9